Keywords: Autonomous Mobility, Demand Prediction, Multi-Output Models, Demand-Responsive Optimization.
Team participants: Inon Peled, Francisco C. Pereira, Carlos Lima Azevedo
Lead Organization: Gate 21
- Prediction of demand for autonomous shuttles in a small area.
- Multi-output models for spatio-temporally correlated movements.
- Interfacing with demand-responsive optimization of supply.
In 2019, autonomous shuttles will start operating in DTU Lyngby campus as part of LINC: the largest test of autonomous shuttles in Denmark. The goal of our PhD project is to dynamically predict where and when passengers would like to use the shuttles for traveling on-campus. To yield such predictions, we develop multi-output models using a variety of machine learning methods, such as Quantile Regression, Deep Learning, and Bayesian Inference. Our project interfaces with a complementary project, conducted in Nanyang Technological University (NTU) in Singapore, which uses our predictions to dynamically optimize the itinerary of the autonomous shuttles.
Sponsors and Partners
- Gate 21
- Roskilde University
- Albertslund Kommune
- Loop City
- Nanyang Technological University (NTU, PhD Program Partner)
- UIA, EU Commission (Sponsors)